1. Field of Invention
The present invention relates to an automatic transmission shift control apparatus and method. In particular, the invention relates to a technology that prevents an overrun of an input rotational speed due to inappropriate learning of a shift point when learning-correction of an upshift point is performed such that the input rotational speed reaches a target maximum rotational speed.
2. Description of the Related Art
A conventional automatic transmission shift control apparatus includes: (a) a shift-point learning-correcting unit that executes learning-correction of an upshift point so that the input rotational speed reaches a predetermined target maximum rotational speed; and (b) a high acceleration upshifting unit that controls, when predetermined high acceleration is requested, the upshift based on a shift point in which a leaning value of the shift point learning-correcting unit is reflected. Japanese Patent Application Publication No. 2004-218799 describes an example of the conventional shift control apparatus. The shift control apparatus described in Japanese Patent Application Publication No. 2004-218799 calculates a virtual maximum rotational speed for a turbine, which is a maximum rotational speed of the turbine when a turbine rotational speed changes at a predetermined reference rotational speed change rate, based on the turbine rotational speed (input rotational speed) when an upshift command is issued, lag time from the issuing of the upshift command to a starting of an inertia phase, and the reference rotational speed change rate. The shift control apparatus then executes, when high acceleration is requested, learning-correction of the shift point (vehicle speed) such that the virtual maximum rotational speed approaches the target maximum rotational speed. Several other conventional shift control apparatuses also perform learning-correction of the shift point directly or indirectly, for example, by using the actual change rate of the turbine rotational speed as a parameter, or by learning the lag time to correct the shift point indirectly.
Even in the shift control apparatuses that perform the learning-correction of the shift point when the high acceleration is requested, the upshift is performed based on a reference value of the shift point in which the learning value of the shift point learning-correcting unit is not reflected when the high acceleration is not requested. However, if the learning-correction of the shift point is performed when the upshift is performed based on the reference value of the shift point, an error will be generated in the learning value. As such, if a mismatch exists between the reflection of the learning value in the shift control and the learning of the shift point, inappropriate learning will occur to cause an overrun of the input rotational speed or a driving power source.